University of South Carolina
In Bayesian theory, the posterior density is proportional to the product of a prior
density and parametric likelihood. This research introduces a new nonparametric
technique that improves upon Bayesian empirical likelihood by replacing the para-
metric likelihood in Bayesian inference with a more robust nonparametric likelihood.
We show that this new technique is a valid likelihood for Bayesian inference. We also
examine frequentist properties of posterior intervals obtained from this methodology.